Breast Cancer Detection and Classification Using a Novel Fast and Robust FCM Segmentation and MWCA based LLRBFNN Machine Learning Model
The breast cancer diagnosis is the main cause of cancer death among women in the world as per the World Health Organization. Breast cancer mortality rates are higher, due to non-availability of early detection especially in underdeveloped and developing countries. This paper proposes of a method to detect and classify mammographic injuries using the regions of attention of breast images. This research work proposes decomposing of each image using a novel FRFCM segmentation and classification of breast cancer categories by utilizing local linear radial basis functional neural network (LLRBFNN). Further to enhance the performance of accuracy of LLRBFNN model, modified water cycle algorithm (MWCA) has been considered to update the weights of LLRBFNN. The results obtained from the proposed MWCA-LLRBFNN model and conventional machine learning models are compared and presented. This approach will combine image and shape of surface features for detection and classification of breast cancer ailments such as benign or malignant.
Keywords: Radial Basis Function Neural Network, Fuzzy C Means Algorithm, Deep
Learning, Water Cycle Algorithm, Convolutional Neural Network